Update README.md
Browse files
README.md
CHANGED
@@ -8,31 +8,24 @@ metrics:
|
|
8 |
model-index:
|
9 |
- name: griffin-1024-llama3t-8layer-simplewiki-silu-fineweb-1M_en-med-vN
|
10 |
results: []
|
|
|
|
|
|
|
|
|
11 |
---
|
12 |
|
13 |
-
|
14 |
-
should probably proofread and complete it, then remove this comment. -->
|
15 |
|
16 |
-
|
17 |
|
18 |
-
|
|
|
|
|
19 |
It achieves the following results on the evaluation set:
|
20 |
- Loss: 5.6538
|
21 |
- Accuracy: 0.1881
|
22 |
- Num Input Tokens Seen: 766509056
|
23 |
|
24 |
-
## Model description
|
25 |
-
|
26 |
-
More information needed
|
27 |
-
|
28 |
-
## Intended uses & limitations
|
29 |
-
|
30 |
-
More information needed
|
31 |
-
|
32 |
-
## Training and evaluation data
|
33 |
-
|
34 |
-
More information needed
|
35 |
-
|
36 |
## Training procedure
|
37 |
|
38 |
### Training hyperparameters
|
@@ -74,4 +67,4 @@ The following hyperparameters were used during training:
|
|
74 |
- Transformers 4.40.1
|
75 |
- Pytorch 2.3.0+cu121
|
76 |
- Datasets 2.19.0
|
77 |
-
- Tokenizers 0.19.1
|
|
|
8 |
model-index:
|
9 |
- name: griffin-1024-llama3t-8layer-simplewiki-silu-fineweb-1M_en-med-vN
|
10 |
results: []
|
11 |
+
datasets:
|
12 |
+
- BEE-spoke-data/fineweb-1M_en-med
|
13 |
+
language:
|
14 |
+
- en
|
15 |
---
|
16 |
|
17 |
+
# griffin-llama3t-8L-v0.02-fineweb
|
|
|
18 |
|
19 |
+
Pretraining experiment with griffin/recurrent_gemma arch. This one uses the Llama-3 tokenizer.
|
20 |
|
21 |
+
## Model description
|
22 |
+
|
23 |
+
Further training of [pszemraj/griffin-1024-llama3t-8layer-simplewiki-silu](https://huggingface.co/pszemraj/griffin-1024-llama3t-8layer-simplewiki-silu) on the BEE-spoke-data/fineweb-1M_en-med dataset.
|
24 |
It achieves the following results on the evaluation set:
|
25 |
- Loss: 5.6538
|
26 |
- Accuracy: 0.1881
|
27 |
- Num Input Tokens Seen: 766509056
|
28 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
## Training procedure
|
30 |
|
31 |
### Training hyperparameters
|
|
|
67 |
- Transformers 4.40.1
|
68 |
- Pytorch 2.3.0+cu121
|
69 |
- Datasets 2.19.0
|
70 |
+
- Tokenizers 0.19.1
|